You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
fix(kimi): export declared effort ladder and default to Kimi Code - #8126
The Kimi Code integration writes one [models."opencodex/<model>"] block per routed model, but buildKimiClientConfig() never read the model's declared effort ladder. A routed model such as kimi/k3.1-pro (reasoningEfforts: ["low","high","max"], defaultReasoningEffort: "low") was exported with no support_efforts, no default_effort and no thinking capability. Kimi Code cannot infer those from a routed name, so it offered no thinking-effort selection even though the proxy accepts reasoning_effort for that model. The DSH exporter already maps the same metadata.
The Kimi block now carries it, following Kimi Code's documented model fields:
support_efforts: the declared ladder, limited to Kimi Code's levels (low, medium, high, xhigh, max) in that order. OpenCodex-only values such as ultra are dropped, because Kimi Code fails validation on an unknown level.
default_effort: only when the declared default is in that exported ladder, since Kimi requires it to be a member.
capabilities gains thinking only when a ladder is exported; image_in is unchanged.
An empty, unknown or entirely unsupported ladder emits none of these fields, so nothing is invented. Context window, display name and image handling are unchanged.
bun test tests/config/client-config-kimi-efforts.test.ts: 6 pass, 0 fail. With the src/ change stashed: 3 pass, 3 fail (the ladder/default, the filtering and the TOML round-trip cases).
bun test tests/config/client-config-kimi-efforts.test.ts tests/config/client-config-new-clients.test.ts tests/config/client-config-export-new-clients.test.ts tests/clients/ tests/ci-workflows/file-size-ratchet.test.ts tests/test-layout.test.ts tests/test-layout-tooling.test.ts: 2572 pass, 0 fail.
tests/server/management-client-config-route.test.ts asserted Kimi capabilities were exactly ["image_in"] for Anthropic models, which declare an effort ladder and now also get thinking. That assertion now checks that image_in is present (52 pass).
bun run typecheck, bun run privacy:scan: pass.
NOT RUN: bun run test:changed and the full bun run test, because the host is shared with several full-suite lanes. Left to CI. Note that test 2/4 is currently red on dev itself: tests/server/server-combo-required-tool-contract.test.ts fails from fix(responses): scope tool authority to current turn #8027 onward, unrelated to this change.
Checklist
Scope stays focused and avoids unrelated cleanup.
Docs or release notes were updated when needed. (The integration docs list only the file path and format, so nothing to update.)
Security-sensitive changes were reviewed for secrets, auth, and unsafe defaults.
Summary by CodeRabbit
New Features
Kimi Code configuration exports now include supported reasoning effort levels and a valid default effort. Thinking and image-input capabilities are included when applicable.
Bug Fixes
Unsupported or unavailable effort levels are omitted from exports, along with defaults that aren’t supported.
The Kimi Code exporter now emits supported reasoning-effort levels and a matching default when declared. It adds the thinking capability when valid effort levels exist and retains image_in when image input is declared.
The exporter filters effort metadata to Kimi’s supported levels and emits a default only when it matches that list. Tests cover filtering, capabilities, and TOML round-tripping. Documentation and test-layout mappings reflect the updated export.
sequenceDiagram
participant ExportModel
participant buildKimiClientConfig
participant TOMLSerialization
ExportModel->>buildKimiClientConfig: declared reasoning efforts and default
buildKimiClientConfig->>buildKimiClientConfig: filter effort levels and build capabilities
buildKimiClientConfig->>TOMLSerialization: generated Kimi config
TOMLSerialization-->>buildKimiClientConfig: serialized TOML
Loading
Merge Risk: Low · previous review 236f0
The change is mergeable with bounded follow-up: malformed hand-edited effort metadata can prevent Kimi configuration export, and the provider guide needs updating.
Pre-merge checks: 4 · 1
Failed checks: 1 warning
Check name
Status
Explanation
Resolution
Docstring Coverage
Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff with cyclomatic complexity of at least 2. Documentat…
Write docstrings for the functions missing them to satisfy the coverage threshold.
Passed checks: 4 passed
Check name
Status
Explanation
Description Check
Check skipped - CodeRabbit’s high-level summary is enabled.
Title check
The title "fix(kimi): export declared effort ladder and default to Kimi Code" matches the main change. The changeset adds support_efforts and default_effort export for Kimi Code model blocks, emits th…
Linked Issues check
Issue #8095 is directly linked and has coding requirements. In src/clients/config-export.ts (the KimiEffort/kimiEfforts code near lines 986-1000), the exporter reads reasoningEfforts and `defa…
Out of Scope Changes check
The changes stay within issue #8095. The source change implements Kimi effort metadata and the related capability. The new focused tests and the management-route assertion update verify the changed be…
Full details: Title check
Explanation
The title "fix(kimi): export declared effort ladder and default to Kimi Code" matches the main change. The changeset adds support_efforts and default_effort export for Kimi Code model blocks, emits thinking capability when a ladder is exported, and adds tests covering these behaviors. The title is specific, concise, and single-sentence, and it names the Kimi scope and the effort-ladder and default behavior.
Full details: Linked Issues check
Explanation
Issue #8095 is directly linked and has coding requirements. In src/clients/config-export.ts (the KimiEffort/kimiEfforts code near lines 986-1000), the exporter reads reasoningEfforts and defaultReasoningEffort, filters to Kimi Code's documented levels, orders and de-duplicates the result, and emits default_effort only when the default is in that result. The build path near lines 1196-1210 adds thinking only when support_efforts exists and preserves the existing context, display-name, and image_in behavior. Empty, undeclared, unsupported-only, and explicitly unsupported reasoning cases emit no effort fields or thinking capability. tests/config/client-config-kimi-efforts.test.ts covers ladder/default export, filtering, ordering, omission cases, capability behavior, and TOML round-trip behavior. The test update in tests/server/management-client-config-route.test.ts preserves the image capability assertion while allowing the new thinking capability.
Full details: Out of Scope Changes check
Explanation
The changes stay within issue #8095. The source change implements Kimi effort metadata and the related capability. The new focused tests and the management-route assertion update verify the changed behavior. The documentation row in structure/clients/integrations.md records the Kimi export contract. The test-layout files register the new test. No unrelated product behavior, context handling, display-name handling, or image handling was changed.
Full details: Docstring Coverage
Explanation
Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff with cyclomatic complexity of at least 2. Documentation-only edits and deletions within surviving functions are included; unchanged diff context and fully deleted functions are excluded. Analyzed 2 functions across 3 files. (3 skipped: 3 unsupported.)
Resolution
Write docstrings for the functions missing them to satisfy the coverage threshold.
Finishing Touches
Generate docstrings
Commit to this branch
Create a new PR
Generate unit tests (beta)
Commit to this branch
Create a new PR
Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.
Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 3 remain after this review.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Summary
The Kimi Code integration writes one
[models."opencodex/<model>"]block per routed model, butbuildKimiClientConfig()never read the model's declared effort ladder. A routed model such askimi/k3.1-pro(reasoningEfforts: ["low","high","max"],defaultReasoningEffort: "low") was exported with nosupport_efforts, nodefault_effortand nothinkingcapability. Kimi Code cannot infer those from a routed name, so it offered no thinking-effort selection even though the proxy acceptsreasoning_effortfor that model. The DSH exporter already maps the same metadata.The Kimi block now carries it, following Kimi Code's documented model fields:
support_efforts: the declared ladder, limited to Kimi Code's levels (low,medium,high,xhigh,max) in that order. OpenCodex-only values such asultraare dropped, because Kimi Code fails validation on an unknown level.default_effort: only when the declared default is in that exported ladder, since Kimi requires it to be a member.capabilitiesgainsthinkingonly when a ladder is exported;image_inis unchanged.Before:
capabilities = ["image_in"], no effort fields.After:
capabilities = ["thinking", "image_in"],support_efforts = ["low", "high", "max"],default_effort = "low".Closes #8095
Verification
bun test tests/config/client-config-kimi-efforts.test.ts: 6 pass, 0 fail. With thesrc/change stashed: 3 pass, 3 fail (the ladder/default, the filtering and the TOML round-trip cases).bun test tests/config/client-config-kimi-efforts.test.ts tests/config/client-config-new-clients.test.ts tests/config/client-config-export-new-clients.test.ts tests/clients/ tests/ci-workflows/file-size-ratchet.test.ts tests/test-layout.test.ts tests/test-layout-tooling.test.ts: 2572 pass, 0 fail.tests/server/management-client-config-route.test.tsasserted Kimi capabilities were exactly["image_in"]for Anthropic models, which declare an effort ladder and now also getthinking. That assertion now checks thatimage_inis present (52 pass).bun run typecheck,bun run privacy:scan: pass.bun run test:changedand the fullbun run test, because the host is shared with several full-suite lanes. Left to CI. Note thattest 2/4is currently red on dev itself:tests/server/server-combo-required-tool-contract.test.tsfails from fix(responses): scope tool authority to current turn #8027 onward, unrelated to this change.Checklist
Summary by CodeRabbit